Naturally Occurring Acute Coliform Mastitis in Holstein Cattle
Christopher K. Cebra, Franklyn B. Garry, and R. Page Dinsmore
Physical examination and clinicopathologic findings from 44
adult Holstein cows with naturally occurring coliform mastitis were studied. The cattle were grouped for comparison
by stage of lactation and survival. Cattle within the first 4
weeks of lactation maintained higher median mature neutrophil counts (1,200 versus 3OO/pL) in peripheral blood than
cattle later in lactation. Nonsurviving cows had higher median creatinine concentration (2.5 versus 1.6 mg/dL) and
anion gap (25 versus 20 mEq/L), and lower serum protein
(7.1 versus 7.6 gm/dL) and total C02 (19.8 versus 25 mEq/L)
concentrations than surviving cows (P< .05). These findings
indicate that cattle with uremia and metabolic acidosis are
less likely t o survive the infection. Bacteriologic blood cultures were performed on 34 of the 44 cows studied. Escherichia coli was isolated from the blood in 11 (32%)cows.
Clinical presentation and clinicopathologic data were compared in bacteremic versus nonbacteremic cows t o evaluate
these data as predictors of bacteremia. Bacteremic cows
were sick longer prior t o admission (2 versus 1 days), maintained higher median counts of total nucleated cells (6.6
versus 2.4 x lo3 cells/pL), myelocytes (0.2 versus 0 x lo3
cells/pL), metamyelocytes (0.5 versus 0.02 x lo3 cells/pL),
band neutrophils (0.7 versus 0.1 x lo3cells/pL), and lymphocytes (2.1 versus 1.4 x lo3 cells/pL) than nonbacteremic
cows, and had higher plasma fibrinogen concentration (600
versus 500 mg/dL) ( P < .05).There were no differences between the physical or serum biochemical measurements.
Four of 11 bacteremic cows and 5 of 23 nonbacteremic cows
died or were euthanized (P> .05). The high prevalence of
bacteremia seen in cows with coliform mastitis has not been
reported previously, and may have been due t o the duration
of disease, severity of signs, or culture technique. These
findings suggest that systemic antibiotic therapy may be
beneficial in some severe cases of coliform mastitis.
J Vet Intern M e d 1996;252-257. Copyright 0 7996 by the
American College of Veterinary lnternal Medicine.
P
Clinicopathologic studies of cows with experimental
ACM have revealed transient abnormalities including leukopenialO.l 1,15,16 consisting of lymphopenia and neutropenia
with left shift,"," hypocalcemia,".'6 and metabolic acidosis." These reports usually deal with a cow's response to a
and the relevance of these
single dose of endotoxin,'0.f',16
abnormalities to the course of the disease in animals with
naturally occurring disease has not been tested. Although
le~kopenia,"'~-'~
high aspartate aminotransferase activ~,~~
hypocalitY>3,18,19 ~ r e m i a , l ~ .h*y~p ~ a l b u m i n e m i a , ~and
~emia'~-''have been reported for small numbers of both
surviving and nonsurviving cows with natural infections, the
importance of these findings relative to disease severity has
not been determined.
In the present study, physical examination and clinicopathologic findings from cows with naturally occurring severe ACM that were admitted to the Colorado State University Veterinary Teaching Hospital (CSU-VTH) were
compared at different stages of lactation and between survivors and nonsurvivors. This report also establishes the existence of bacteremic ACM, and demonstrates that this clinical
syndrome is very similar to nonbacteremic ACM.
eracute and severe acute coliform mastitis (ACM) are
combinations of local and systemic inflammatory disease, and result from invasion of the mammary gland by
Escherichia coli or a similar coliform organism.'-3 Cattle in
early lactation are reported to be more susceptible to the
d i ~ e a s e .This
~ - ~ may be attributable to decreased neutrophil
function,7.*delayed migration of neutrophils into the gland,6,9
and faster bacterial replication during early l a ~ t a t i o nFailure
.~
to rapidly clear the organism may allow the bacteria to multiply and release more endotoxin prior to the onset of the
inflammatory reaction.
The local and systemic signs resemble the reaction to
endotoxin, and may be induced by intramammary administration of endotoxin. "," Furthermore, bacteremia has not
been demonstrated in cattle with ACM." Because of this,
current therapeutic recommendations are centered around
removal and inhibition of the toxin and not elimination of the
bacteria with antimicrobial agents."," Lack of attachment
to epithelial cells and maintenance of basement membrane
integrity have been seen ultrastructurally after experimental
infection, and were presumed to prevent systemic invasion
by the o r g a n i ~ m .In
' ~ contrast, some destruction of the basement membrane of mammary tissue has been seen in early
lactation cows with naturally occurring ACM. Bacteria were
isolated from regional lymph nodes of these cattle.'
From the Department of Clinical Sciences, College of Veterinary
Medicine and Biomedical Sciences, Colorado State University, Fort
Collins, CO.
Accepted December 29, 1995.
An abstract of this study was presented at the 12th Annual ACVIM
Forum.
Supported in part by Colorado State University Agricultural Experiment Station Funds.
Reprint requests: Christopher K . Cebru, VMD, Department of
Clinical Sciences, College of Veterinary Medicine, Colorado State
University, Fort Collins, CO 80523.
Copyright 0 I996 by the American College of Veterinary Internal
Medicine
0891-6640/96/I004-00O9$3.00/0
252
Materials and Methods
Case records were reviewed for 44 adult Holstein cows with ACM
admitted to the CSU-VTH between January 1, 1990 and February
28, 1995. Only cattle with a record of historical information, physical
examination findings, milk culture positive for E coli, and a CBC
performed at admission were selected. Cows with chronic disease
and those not demonstrating systemic signs were excluded. Serum
chemistry values, blood gas analyses, and noncontarninated blood
bacteriologic culture results were used when available. These 44
cows represent 61% of the Holstein cattle that were admitted for
mastitis with systemic signs. Reasons for exclusion of the other 28
animals included no milk culture (n = 15), isolation of exclusively
Gram-positive organisms (n = 4), or isolation of Gram-negative
organisms without E coli (n = 9). Other Gram-negative isolates
included Klebsiella sp. (n = 3), Pseudomonas sp. (n = 2), Serratia
sp. (n = 2), Salmonella sp. (n = l), and Pasteurella sp. (n = 1).
The decision whether to obtain a milk sample for culture was more
Journal of Veterinary lnternal Medicine, Vol 10, No 4 (July-August), 1996: pp 252-257
COLIFORM MASTITIS IN HOLSTEIN COWS
dependent on the attending clinician than the status of the animals.
The 9 cows with gram-negative isolates other than E coli were
excluded because they increased the heterogeneity of the study population.
Blood samples for clinical pathology testing were collected aseptically from either the jugular or tail veins. Automated cell counts
(Coulter S+ IV; Coulter Corp, Hialeah, E),serum chemistry profiles (Hitachi 91 1 Automatic Analyzer; Boehringer Mannheim, Indianapolis, IN), and blood gas analysis (ABL 300 Acid-Base Analyzer;
Radiometer, West Lake, OH) were performed. Milk culture samples
were collected into sterile tubes after the teat end was cleaned with
isopropyl alcohol, and streaked onto Columbia and MacConkey agar
plates. The plates were incubated at 35" to 37°C in a 10% COz
atmosphere. Colonies were identified by standard laboratory methods. Sensitivity to antimicrobial agents was determined by minimum
inhibitory concentration using commercial plates (Sensititre
CBV6LG; Radiometer, West Lake, OH).
For blood culture, the hair was clipped from a site over the jugular
vein. The site was then disinfected by alternating applications of
povidone iodine and isopropyl alcohol. Twenty milliliters of blood
were removed aseptically from the vein and injected aseptically
through a new needle into a 50-mL blood culture vial of Brain
Heart infusion containing 0.6% sodium polyanetholsulfonate (BBL;
Becton Dickinson Co, Cockeysville, MD). The sample was aerated
through the needle after blood inoculation, incubated, and subcultured as described above on days 0, 1 and 7.
For comparison, the cattle were stratified in 3 separate groups.
For the first comparison, cattle were grouped by outcome: the survivor group (S) consisted of cattle that were discharged and the nonsurvivor group (NS) of cattle that died or were euthanized during their
hospital stay. Euthanasia was considered to be an identical outcome
to natural death because cattle were only euthanized for humane
reasons when their condition failed to improve with therapy. Such
cattle were characteristically unable to rise, tachycardic, completely
anorectic, and often in respiratory distress.
For the second comparison, the cattle were divided into 3 groups
based on their stage of lactation. The 3 stages were early (EL; 1 to
28 days), middle (ML; 29 to 119 days), and late (LL; 120 or more
days) lactation.
For the third comparison, a noncontaminated blood culture sample
was necessary for inclusion. Thirty-four of the 44 cows fulfilled this
criterion. These cattle were classified as bacteremic or nonbacteremic
based on the presence or absence of E coli in the blood culture.
Parameters used for comparison consisted of historical, physical
examination, and clinicopathologic data. Specifically, age in years,
days in milk, days sick prior to admission, prior treatment, rectal
temperature, heart rate, and respiratory rate were taken from reports
of the initial examination. Laboratory samples were taken at admission to the CSU-VTH. PCV and total and differential leukocyte
counts were taken from the CBC. Serum urea nitrogen, glucose,
creatinine, calcium, phosphorus, total protein, albumin, globulin,
albumin-globulin ratio, total bilirubin, creatine kinase (CK) activity,
aspartate aminotransferase (AST) activity, sodium, potassium, chloride, total C 0 2 , and anion gap were taken from the serum chemistry
profile. Blood pH and bicarbonate were measured by blood gas
analysis.
For statistical analysis of 2-group comparisons, the Wilcoxon
rank-sum test" was used if the data were continuous and Fischer's
exact testz2was used if the data were ordinal. The data from the 3
groups of cows stratified by days in milk were compared by the
Kruskal-Wallis test?' The Kruskal-Wallis tests and internal comparisons between groups were performed using Sigmastat statistical
software (Jandel Scientific Software, San Rafael, CA). Statistical
significance was determined when P < .05.
253
Results
In addition to E coli, bacteria isolated from milk samples
included Staphylococcus sp. (n = 5 ) , Streptococcus sp. (n
= l), and Klebsiella sp. (n = 1). Two or 3 colony types of
E coli were isolated from 7 cows. Although sensitivity to
tetracycline was inconsistent (54%), bacterial colonies resistant to gentamicin and ceftiofur were isolated only from 1
cow each. Prior to admission, 26 of 44 cows had been treated
by the owners or herd veterinarians with an intramammary
infusion of either gentamicin (n = 17) or cephapirin sodium
(n = 8), or both (n = 1). T w o cows treated with intramammary gentamicin were also treated with intramammary polymixin B. Nineteen of the 26 cattle that had received intramammary therapy also had been treated with 1 or more
systemic antibiotic agent (oxytetracycline [n = 81, ceftiofur
sodium [n = 31, procaine penicillin G [n = 51, gentamicin
[n = 21, or sulfa-based antibiotics and macrolides [n = 4]),
and 4 additional cows had been treated with systemic antibiotics alone (ceftiofur sodium [n = 21, gentamicin [n = I],
or oxytetracycline [n = 11). Fourteen cows had n o history of
antibiotic treatment; 22 cows had been treated with flunixin
meglumine (Banamine; Schering-Plough Animal Health,
Kenilworth, NJ).
The historical, CBC, serum chemistry, and blood gas analysis values for the study population are listed in Tables 1
and 2. Cattle had been sick for a median of 2 days (range,
1 to 7 days). Lactating cattle of all ages and stages of lactation were affected. Characteristic findings o n physical examination were heart rates above 80 beats per minute (n = 39)
and respiratory rates above 28 breaths per minute (n = 31).
Nine had rectal temperatures above 39.5"C and 9 had temperatures below 37.5"C. Th e leukogram characteristically
showed leukopenia, consisting of lymphopenia (n = 33),
neutropenia (n = 26), and a left shift of the neutrophil line
(n = 36). The serum chemistry analysis showed increases
in blood urea nitrogen (21 of 23), creatinine (19 of 38),
Table 1. Historical, Physical, and Hematologic
Findings in 44 Cows With Naturally Occurring Acute
Coliform Mastitis
Variable
Age (y)
Days in milk
Days sick
Rectal Temperature PC)
Heart rate (beathin)
Respiratory rate (breathhin)
Nucleated Cells (x103/pL)
Myelocytes (x103/pL)
Metamyelocytes ( x 103/pL)
Band neutrophils (x103/pL)
Mature neutrophils (xIO3/pL)
Lym p hocytes (x103/pL)
PCV (%)
Fibrinogen (mg/dL)
Median
(min-max)
5 (2-12)
65 (1-300)
2 (1-7)
38.7 (35.3-40.6)
110 (66-150)
40 (20-88)
3.6 (0.7-27.5)
0 (0-2.8)
0.1 (0-6.9)
0.2 (0-13.8)
0.4 (0-5.0)
1.7 (0.5-8.7)
36 (27-48)
600 (200-1,100)
Reference
Values*
4-12
0
0
0-0.1
0.6-4.0
2.5-7.5
24-46
200-400
* Reference values used by the Colorado State University Clinical
Pathology Laboratory.
254
CEBRA, GARRY, AND DINSMORE
Table 2. Serum Chemistry and Blood Gas Analysis
Results in 44 Cows With Naturally Occurring
Acute Coliform Mastitis
Variable
Serum urea nitrogen (mg/dL)
Glucose (mg/dL)
Creatin i ne ( mg/d L)
Calcium (mg/dL)
Phosphorus (mg/dL)
Total protein (g/dL)
Albumin (g/dL)
Globulin (g/dL)
Albumin/globulin
Bilirubin (mg/dL)
Creatine phosphokinase (IU/L)
Aspartate aminotransferase (IU/L)
Sodium (mEq/L)
Potassium (mEq/L)
Chloride (mEq/L)
Total C02 (mEq/L)
Anion gap (mEq/L)
Blood pH
Bicarbonate (mEq/L))
Median
(min-max)
36 (13-89)
84 (38-164)
1.7 (0.9-4.0)
8.2 (5.1-28.8)
8.3 (4.1-12.9)
6.9 (5.7-8.3)
3.4 (2.7-4.6)
3.3 (2.1-5.5)
1 (0.5-2.0)
0.7 (0.2-6.6)
506 (86-55,580)
108 (43-873)
138 (130-152)
4.0 (2.4-5.7)
97 (85-115)
24.1 (12.7-31.5)
21 (12-35)
7.46 (7.24-7.53)
25.8 (9.3-32.4)
Reference
Values’
7-20
55-95
0.9-1.7
7.6-10.2
4-8.6
6.4-9.5
2.5-4.3
2.6-6.5
0.5-1.3
0.1-0.4
57-280
40-130
136-147
4.0-5.0
95-105
21 -27
14-26
7.31-7.53
25-35
* Reference values used by the Colorado State University Clinical
Pathology Laboratory.
phosphorus (16 of 37), CK (28 of 37), AST activity (15 of
37), anion gap (8 of 37), and bilirubin (31 of 37) and decreases in potassium (18 of 39), calcium (10 of 37), and
chloride (14 of 38). Measurement of systemic acid-base balance revealed polarized results: total C 0 2 >27 mEq/L (9 of
37), pH >7.45 (14 of 27), and bicarbonate >24 mEq/L (17
of 27) were common, but total CO, <21 mEq/L (9 of 37),
pH < 7.35 (5 of 27), and bicarbonate < I8 mEq/L (5 of 27)
were also encountered.
All cattle were treated for ACM at the clinic. Standard
treatment consisted of IV polyionic replacement fluids (40
to 140 L), flunixin meglumine (0.25 to 0.5 mgkg, IV, every
6 to 8 hours), and stripping out affected quarters every 2
hours. Additionally, 41 of 44 cows were treated with antibiotics systemically (ceftiofur sodium [n = 361, procaine penicillin G [n = 181, others [n = 41) or intramammary (gentamicin [n = 1.51, cephapirin sodium [n = 141, cloxacillin [n =
11). Three cows were not treated with antibiotics at the owners’ request. Twelve of the 44 cows died or were euthanized.
Comparison by Outcome
Comparison of S to NS cows (Table 3) showed nonsurvivors had significantly higher serum creatinine, sodium, potassium, AST activity, and anion gap, and significantly lower
serum protein, total C 0 2 , and blood pH ( P < .05). However,
the protein and anion gaps for both groups were most commonly within normal reference ranges. Sodium, potassium,
and chloride concentrations were more likely to be abnormally low among survivors than high among nonsurvivors,
and total C 0 2 , blood pH, and bicarbonate were more likely
to be abnormally high in survivors than in nonsurvivors.
Sodium was characteristic of the electrolytes, in that NS
cows tended toward higher values over a similar range as S
cows (Fig l), while total CO, was typical of measurements
of metabolic acid-base status, in that S cows rarely were
acidemic, whereas NS cows commonly were (Fig 1). Seven
of 8 cows with a serum total C 0 2 <20 m E q L died or were
euthanized, and 3 of the 4 nonacidemic NS cows were both
alkalemic and bacteremic. Prior antibiotic therapy was similar in survivors (24 of 32) and nonsurvivors (8 of 12).
Comparison by Stage of Lactation
There were 12 cows in early lactation (EL), 20 in middle
lactation (ML), ans 12 in late lactation (LL). There were
significant differences among cattle at different stages of
lactation for PCV and peripheral blood counts of mature
neutrophils and lymphocytes ( P < .01), and the difference
in peripheral counts of nucleated cells ( P = .12) and band
neutrophils ( P = .08) approached significance (Table 4). The
EL cows maintained higher median numbers of band cells
and neutrophils than the other 2 groups. The LL cows maintained a higher PCV than either other group. Five of 12 EL
cattle, 3 of 20 in the ML group, and 4 of 12 in the LL group
died or were euthanized.
Comparison by Blood Culture Results
Blood culture results were available from single samples
on 33 cows, and from paired samples on 1 cow, and were
positive for E coli from 11 cows (32%). An additional bacterial isolate was found for 2 bacteremic cows (one each of
Staphylococcus and Pasteurella). The pattern of antibiotic
sensitivity of the blood isolate was very similar or identical
to that for a mammary isolate. Previous antibiotic therapy
was more common in bacteremic than nonbacteremic cows
(72% versus 61%), but a smaller percentage of bacteremic
cows (36% versus 48%) had received systemic antibiotic
Table 3. Comparison of Selected Hematologic and
Serum Chemistry Values for Surviving and
Nonsurviving Cows With Naturally Occurring
Acute Coliform Mastitis
Survivors
( n = 32)
Variable
PCV (Yo)
Creatinine (mg/dL)
Total protein (g/dL)
Globulin (g/dL)
Aspartate
aminotransferase
(IUIL)
Sodium (mEq/L)
Potassium (mEq/L)
Chloride (mEq/L)
Total CO? (rnEq/L)
Anion gap (mEq/L)
Blood pH
Bicarbonate (mEq/L)
Median
(min-max)
35 (27-46)
1.6(1.0-4.0)
7.6 (5.7-8.3)
3.5 (2.4-5.5)
105 (48-4391
137 (131-148)
3.9 (2.4-5.2)
96 (86-110)
25 (19.1-31.5)
20 (12-28)
7.47(7.26-7.53)
25.9 (9.3-30.9)
Non-Survivors
(n = 12)
PValue
.13
,002
.01
.15
,005
.02
.03
.08
.03
,005
,003
.05
Median
(min-max)
39 (32-48)
2.5 (1.7-3.9)
7.1 (5.7-7.0)
3.3 (2.8-3.6)
187 (84-873)
145 (134-152)
4.7 (2.6-5.7)
101 (85-115)
19.8 (12.7-30.4)
25 (20-35)
7.35 (7.24-7.41)
19.7 (9.7-29.6)
COLIFORM MASTITIS IN HOLSTEIN COWS
255
153
s
2 145
-E
G
a
R
E
2 137
4:
129
N-S
A
S
Outcome
32
4
12
i
0
I
N- S
B
S
Outcome
Fig 1. Box and whisker plots representing the serum concentrations of (A) sodium and (B) total carbon dioxide in nonsurviving INS) and surviving (SIcows with acute coliform mastitis. The central
box contains the interquartile range, with the vertical bars extending
t o the maximum and minimum values. The median is represented
by the horizontal bar, which divides the interquartile range.
therapy. The prior use of antibiotics was not significantly
different between bacteremic and nonbacteremic cows (P>
.05).
Bacteremic cows maintained significantly higher fibrinogen concentrations and counts of all forms of immature neutrophils and lymphocytes than did nonbacteremic cows (P
< .05) (Table 5 ) . Although a left shift was common in
both groups of cows, large numbers of immature neutrophils
rarely were found in nonbacteremic animals. Other physical
parameters and clinicopathologic data from bacteremic cows
were similar to those from nonbacteremic cows. Survival
did not appear to relate to stage of lactation and was similar
in bacteremic cows (7 of 11) and nonbacteremic cows (18
of 23). The duration of illness prior to admission was longer
for the bacteremic versus nonbacteremic cows ( P = .01).
Discussion
The clinicopathologic abnormalities seen in this study
population resembled those reported for cows with both ex-
perimentallO.l 1.15.16 and natural coliform m a ~ t i t i s . ~ *HowI~-~'
ever, although the noted changes were transient (less than
24 hours) with experimental'0r'1,15.'6or mild natural infection," they were found several days after the onset of signs
in our studied cattle. This may represent a persistence or
recurrence of these changes.
Lymphopenia and neutropenia with left shift were fairly
ubiquitous changes, and thus of no value in predicting outcome. There was an association between the magnitude of
the left shift and bacteremia, such that suspicion of bacteremia should rise if high counts of myelocytes, metamyelocytes, or band neutrophils are detected. It may be that this
left shift is part of the normal regenerative response, because
bacteremic cows had a longer history of illness prior to
admission. Neutropenia was not a consistent finding in early
lactation cows. Impaired neutrophil mobilization and delayed diapedesis into the gland also have been described in
early postparturient
with the degree of impairment
related to the severity of the disease.24The higher neutrophil
counts seen in early lactation cows in this study may be
additional evidence of decreased responsiveness to coliform
mammary infection in early lactation cows, although increased release from bone marrow may not be excluded. In
spite of this consideration, nonsurvivors occurred with similar frequency in early and late lactation cows.
In contrast to blood cellular changes, biochemical abnormalities were associated with outcome. Hemoconcentration
(high PCV), uremia, high AST activity, and organic acidosis,
all evidence of systemic dysfunction, were typical of nonsurviving cows, whereas decreased concentrations of sodium,
potassium, and chloride, and metabolic alkalosis were often
encountered in surviving cows. The trend towards higher
PCV and lower serum protein concentration in nonsurviving
cows may relect hernoconcentration in the face of protein
loss into the mastitic
or gastrointestinal tract.
The pattern of acid-base abnormalities was similar to that
reported for cows affected with abomasal V O ~ V U ~More
US.~~~~~
severely affected cows developed organic acidosis with increasing anion gap, due to deteriorating cardiovascular function, whereas survivors had a tendency towards lower serum
Table 4. Comparison of Hematologic Values for 44
Cows With Naturally Occurring Acute Coliform
Mastitis at Different Stages of Lactation
Variable
~
Early
( n = 12)
Middle
(n = 20)
Late
( n = 12)
Median
(min-max)
Median
(min-rnax)
Median
(min-rnax)
~
Days in Milk
7 (1-27)
Band cells (x103/pL) 0.4 (0-13.8)
Mature neutrophils
( X 1031p~)
1.2' (0.1-5.0)
Lymphocytes
(x 103/@~)
2.2" (1.1-4.4)
PCV (%)
35" (28-40)
~
65 (30-114)
0.2 (0-3.0)
~
216 (120-300)
0.2 (0-0.7)
0.5@(0-2.5)
0.2b (0-0.7)
1.5b (0.5-4.3)
34a 127-48)
3.6' (1.3-8.7)
40b (34-458)
NOTE. Values marked with different superscripts are significantly
different ( P i .05).
CEBRA, GARRY, AND DINSMORE
256
Table 5. Comparison of Selected Hematologic Values
for Bacteremic and Nonbacteremic Cows With
Naturallv Occurring Acute Coliform Mastitis
Nonbacteremic
Variable
Days sick
Rectal temperature
(C”)
Nucleated cells
(x103jg~)
Myelocytes (x103//1L)
Metamyelocytes
( x 1o 3 i V ~ )
Band neutrophils
( X 1O
~/~L)
Median
(rnin-max)
l(1-7)
Bacteremic
PValue
.01
Median
(rnin-max)
2 (1-4)
.I8
38.4 (35.3-40.1)
2.4 (0.7-9.3)
0 (0-0.1)
.01
.01
6.6 (1.6-27.5)
0.2 (0-2.8)
0 (0-0.3)
.002
0.5 (0-6.9)
0.1 (0-1.1)
,003
0.7 (0.1-13.8)
0.4 (0-3.4)
.46
.03
,008
0.5 (0.1-5.0)
2.1 (0.7-8.5)
600 (400-1.100)
38.7 (36.1-40.5)
Mature neutrophils
(X~O~/~L)
Lymphocytes (x103//1L) 1.4 (0.5-8.6)
500 (200-1,100)
Fibrinogen (mg/dL)
chloride concentration and alkalosis, likely due to endotoxin
induced gastrointestinal stasis and third-space sequestration
of abomasal secretion. In cows with ACM, this trend of
association between acidosis and nonsurvival may have a
similar basis as in cows with abomasal volvulus. Data from
the cows in this study support the conclusion that, in spite
of the diagnostic and therapeutic modalities available in the
referral hospital setting, cows with ACM and metabolic acidosis and uremia are unlikely to survive.
The high prevalence of bacteremia (32%) in these cows
was contrary to previous studies,12 and was perhaps the most
important finding in this study relative to treatment considerations. The study population consisted of cows referred for
treatment, often after on-farm therapy had failed. These cows
had severe, protracted disease, and are not representative of
the overall population of cows with ACM. In contrast, a
previous study that did not detect bacteremia was performed
on field cases with early signs of the disease.12 Thus, it
appears reasonable to presume that cows with ACM and
minimal systemic signs are unlikely to be bacteremic.
Culture technique may have been important in the detection of bacteremia. Culture samples consisted of a minimum
of 20 mL of blood, with subcultures plated out on 3 separate
occasions, compared with 5-mL blood samples and 1 to 2
subcultures used in another study.” It has been shown that
success at culturing bacteria from the blood relates to the
volume of the sample.zxRepeated sampling and greater dilution of the sample in culture media, which helps to neutralize
antibiotic effects, might have led to even greater yields.
Because these cattle had naturally occurring disease, it is
impossible to determine whether uncontrolled factors may
have led to the concurrent mastitis and bacteremia. The
mammary and blood E coli isolates may have been from
unrelated foci of infection, the blood-borne bacteria may
have seeded the mammary gland, or another disease process
in the body may have aided the bacteria’s invasion. Evidence
of hematogenous coliform mastitis has been reported,’,’’ although it is unlikely to be the cause of most coliform intramammary infections. Because of the similarities in antibiotic
sensitivities determined on the milk and blood-borne organisms, it is unlikely that coincidental infections were the
cause. Polymicrobial infections are uncommonly reported in
cows, yet several of the cows in this study had multiple
intramammary pathogens, one also had concurrent Pasteurellu bacteremia and another Stuphyloccocal bacteremia.
These additional organisms may have aided in the damage
to the basement membrane of the mammary gland.
In conclusion, many of the transient clinicopathologic abnormalities noted in cows with experimental ACM were
persistent or recurrent in cows with naturally occurring disease. Hematologic data had greater correlation to bacteremia,
whereas biochemical data were more related to outcome.
Cows with large numbers of immature neutrophils in peripheral blood were often bacteremic, and bacteremic cows as
well as those with biochemical evidence of metabolic derangement often did not survive the disease. Although the
efficacy of different treatment protocols was not evaluated
in this study, the occurrence of bacteremia in some cows
with severe or protracted ACM supports a role for systemic
antibiotic therapy in the treatment of this disease.
References
1. Schalm OW, Woods GM. Characteristics of coliform mastitis
and treatment with dihydrostreptomycin. J Am Vet Med Assoc 1952;
1201385-388.
2. Eberhatt RJ. Coliform mastitis. J Am Vet Med Assoc 1977;
170:1160- 1163.
3. Radostits OM. The clinical aspects of coliform mastitis in cattle. Proceedings of the VI International Conference on Cattle Diseases. Stillwater, OK: Heritage Press; 1970:67-74.
4. Hill AW, Shears AL, Hibbitt KG. The elimination of serumresistant Escherichia coli from experimentally infected single mammary glands of healthy cows. Res Vet Sci 1978;25:89-93.
5. Shuster DE, Kehrli ME. Comparison of Escherichia coli mastitis in early lactation and mid-lactation dairy cows. J Dairy Sci 1993;
76:159 (suppl 1).
6. Hill AW, Shears AL, Hibbitt KG. The pathogenesis of experimental Escherichia coli mastitis in newly calved dairy cows. Res
Vet Sci 1979;26:97-101.
7. Kehrli ME, Nonnecke BJ, Roth JA. Alteration in bovine neutrophil function during the periparturient period. Am J Vet Res 1989;
501207-2 14.
8. Cai T, Weston PG, Lund LA, et al. Association between neutrophil functions and periparturient disorders in cows. Am J Vet Res
1994;55:934-943.
9. Frost AJ, Brooker BE. Hyperacute Escherichia coli mastitis of
cattle in the immediate post-partum period. Aust Vet J 1986;63:
327-331.
10. Carroll EJ, Schalm OW, Lasmanis J. Experimental coliform
(Aerobrrcrer aerogenes) mastitis: Characteristics of the endotoxin
and its role in pathogenesis. Am J Vet Res 1964;25:720-726.
11. DeGraves FJ, Anderson KL. Ibuprofen treatment of endotoxin-induced mastitis in cows. Am J Vet Res 1993;54:1128-1132.
12. Powers MS, White ME, Dinsmore P, et al. Aerobic blood
culturing in cows with coliform mastitis. J Am Vet Med Assoc 1986;
189:440-441.
13. Erskine RJ, Tyler JW, Riddell MG, et al. Theory, use, and
COLIFORM MASTITIS IN HOLSTEIN COWS
realities of efficacy and food safety of antimicrobial treatment of
acute coliform mastitis. J Am Vet Med Assoc 1991; 198:980-984.
14. Frost AJ, Hill AW, Brooker BE. Pathogenesis of experimental
bovine mastitis following a small inoculum of Escherichia coli. Res
Vet Sci 1982;33:105-112.
15. Schalm OW, Lasmanis J, Carroll EJ. Pathogenesis of experimental coliform (Aerobacter aerogenes) mastitis in cattle. Am J Vet
Res 1964;25:75-82.
16. Griel LC, Zarkower A, Eberhart RJ. Clinical and Clinicopathological effects of Escherichia coli endotoxin in mature cattle. Can
J Comp Med 1975;39:1-6.
17. Vestweber JGE, Kruckenberg SM, Spencer H, et al. Disseminated intravascular coagulation in a cow with coliform mastitis.
Comp Cont Educ Pract Vet 1983;5:S185-S190.
18. DeJong GA. Acute mastitis in cows-Part I: Fatal cases.
Mod Vet Pract 1987;68:430-435.
19. Oetzel GR. Coliform mastitis and hypocalcemia in two dairy
cows in mid lactation. Comp Cont Educ Pract Vet 1985;7:S237S242,S244.
20. Katholm J, Haubro Andersen P. Acute coliform mastitis in
dairy cows: Endotoxin and biochemical changes in plasma and colony-forming units in milk. Vet Rec 1992; 131:513-514.
21. Milton JS, Tsokos JO. Statistical Methods in the Biological
and Health Sciences. New York, NY: MacGraw-Hill; 1983:429433.
257
22. Matthews DE, Farewell VT. Using and Understanding Medical Statistics. Basel, Switzerland: Karger; 1985;20-38.
23. Milton JS, Tsokos JO. Statistical Methods in the Biological
and Health Sciences. New York, NY: MacGraw-Hill; 1983:433438.
24. Kremer WDJ, Noordhuizen-Stassen EN, Grommers FJ, et
al. Preinfection chemotactic response of blood polymorphonuclear
leukocytes to predict severity of Escherichia coli mastitis. J Dairy
Sci 1993;76:1568-1574.
25. Shuster DE, Kehrli ME. Pathophysiology, host defense, and
cytokine production during coliform mastitis in midlactation dairy
cows. J Dairy Sci 1992;75:160 (suppl 1).
26. Garry F3, Hull BL, Rings DM, et al. Prognostic value of
anion gap calculation in cattle with abomasal volvulus: 58 cases
(1980-1985). J Am Vet Med Assoc 1988;192:1107-1112.
27. Simpson DF, Erb HN, Smith DF. Base excess as a prognostic
and diagnostic indicator in cows with abomasal volvulus or
right displacement of the abomasum. Am J Vet Res 1985;46:796797.
28. Washington JA 11. Conventional approaches to blood culture.
In: Washington JA 11, ed. The Detection of Septicemia. West Palm
Beach, FL: CRC Press; 1978;44-46.
29. Heidrich HJ, Renk W. Diseases of the Mammary Glands of
Domestic Animals. Philadelphia, PA: WB Saunders; 1967;213.